Dear Google Team, I am writing to respectfully request the inclusion of canonical Pali texts, specifically the Digha Nikaya, Majjhima Nikaya, Samyutta Nikaya, and Anguttara Nikaya, and Kuddaka Nikaya up to Thera Theri Gatā into your datasets. As these are foundational texts of widely studied Buddhism, their accurate and comprehensive inclusion would significantly enhance the performance and accuracy of Gemini models in processing related queries. Thank you for your consideration."
1 Like
Hey,
Hope you’re keeping well.
Gemini’s training data sources aren’t something that can be directly modified or extended by user request. If you need the models to work with specific canonical Pali texts, the practical approach is to provide them as part of a custom tuning or grounding workflow in AI Studio. You can upload these texts as a dataset in Google Cloud Storage, then use fine-tuning or embedding models in Vertex AI to adapt Gemini’s responses to your domain. For retrieval-augmented generation, store the processed texts in BigQuery or a vector database via Vertex AI Search, and connect that to your prompt pipeline so the model always references your authoritative content.
Thanks and regards,
Taz